NVIDIA Maxwell兼容性指南 Release 12.3说明书
合集下载
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
▶ Download and install the latest driver from https:///drivers. ▶ Set the environment variable CUDA_FORCE_PTX_JIT=1. ▶ Launch your application. When starting a CUDA application for the first time with the above environment flag, the CUDA driver will JIT-compile the PTX for each CUDA kernel that is used into native cubin code. If you set the environment variable above and then launch your program and it works properly, then you have successfully verified Maxwell compatibility. Note: Be sure to unset the CUDA_FORCE_PTX_JIT environment variable when you are done testing.
2
Contents
Chapter 1. About this Document
This application note, Maxwell Compatibility Guide for CUDA Applications, is intended to help developers ensure that their NVIDIA® CUDA® applications will run on GPUs based on the NVIDIA® Maxwell Architecture. This document provides guidance to developers who are already familiar with programming in CUDA C++ and want to make sure that their software applications are compatible with Maxwell.
Maxwell Compatibility Guide
Release 12.3 NVIDIA
Nov 14, 2023
Contents
1 About this Document
3
ቤተ መጻሕፍቲ ባይዱ
2 Application Compatibility on Maxwell
5
3 Verifying Maxwell Compatibility for Existing Applications
3.1. Applications Using CUDA Toolkit 5.5 or Earlier
CUDA applications built using CUDA Toolkit versions 2.1 through 5.5 are compatible with Maxwell as long as they are built to include PTX versions of their kernels. To test that PTX JIT is working for your application, you can do the following:
5
Maxwell Compatibility Guide, Release 12.3
6
Chapter 2. Application Compatibility on Maxwell
Chapter 3. Verifying Maxwell Compatibility for Existing Applications
Chapter 4. Building Applications with Maxwell Support
When a CUDA application launches a kernel, the CUDA Runtime determines the compute capability of each GPU in the system and uses this information to automatically find the best matching cubin or PTX version of the kernel that is available. If a cubin file supporting the architecture of the target GPU is available, it is used; otherwise, the CUDA Runtime will load the PTX and JIT-compile that PTX to the GPU’s native cubin format before launching it. If neither is available, then the kernel launch will fail. The method used to build your application with either native cubin or at least PTX support for Maxwell depend on the version of the CUDA Toolkit used. The main advantages of providing native cubins are as follows:
1 Future CUDA Toolkit version might deprecate support for the Maxwell Architecture.
8
Chapter 3. Verifying Maxwell Compatibility for Existing Applications
7
3.1 Applications Using CUDA Toolkit 5.5 or Earlier . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 Applications Using CUDA Toolkit 6.0 or Later . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Maxwell Compatibility Guide for CUDA Applications The guide to building CUDA applications for GPUs based on the NVIDIA Maxwell Architecture.
Contents
1
Maxwell Compatibility Guide, Release 12.3
The first step is to check that Maxwell-compatible device code (at least PTX) is compiled in to the application. The following sections show how to accomplish this for applications built with different CUDA Toolkit versions.
3
Maxwell Compatibility Guide, Release 12.3
4
Chapter 1. About this Document
Chapter 2. Application Compatibility on Maxwell
The NVIDIA CUDA C++ compiler, nvcc, can be used to generate both architecture-specific cubin files and forward-compatible PTX versions of each kernel. Each cubin file targets a specific computecapability version and is forward-compatible only with GPU architectures of the same major version number. For example, cubin files that target compute capability 3.0 are supported on all computecapability 3.x (Kepler) devices but are not supported on compute-capability 5.x (Maxwell) devices. For this reason, to ensure forward compatibility with GPU architectures introduced after the application has been released, it is recommended that all applications include PTX versions of their kernels. Note: CUDA Runtime applications containing both cubin and PTX code for a given architecture will automatically use the cubin by default, keeping the PTX path strictly for forward-compatibility purposes. Applications that already include PTX versions of their kernels should work as-is on Maxwell-based GPUs. Applications that only support specific GPU architectures via cubin files, however, will need to be updated to provide Maxwell-compatible PTX or cubins.
4 Building Applications with Maxwell Support
9
4.1 Applications Using CUDA Toolkit 5.5 or Earlier . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2 Applications Using CUDA Toolkit 6.0 or Later . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
7
Maxwell Compatibility Guide, Release 12.3
3.2. Applications Using CUDA Toolkit 6.0 or Later
CUDA applications built using CUDA Toolkit 6.0 or Later1 are compatible with Maxwell as long as they are built to include kernels in either Maxwell-native cubin format (see Building Applications with Maxwell Support) or PTX format (see Applications Using CUDA Toolkit 5.5 or Earlier) or both.
6.3 Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
i
ii
Maxwell Compatibility Guide, Release 12.3
5 Revision History
13
6 Notices
15
6.1 Notice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2 OpenCL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2
Contents
Chapter 1. About this Document
This application note, Maxwell Compatibility Guide for CUDA Applications, is intended to help developers ensure that their NVIDIA® CUDA® applications will run on GPUs based on the NVIDIA® Maxwell Architecture. This document provides guidance to developers who are already familiar with programming in CUDA C++ and want to make sure that their software applications are compatible with Maxwell.
Maxwell Compatibility Guide
Release 12.3 NVIDIA
Nov 14, 2023
Contents
1 About this Document
3
ቤተ መጻሕፍቲ ባይዱ
2 Application Compatibility on Maxwell
5
3 Verifying Maxwell Compatibility for Existing Applications
3.1. Applications Using CUDA Toolkit 5.5 or Earlier
CUDA applications built using CUDA Toolkit versions 2.1 through 5.5 are compatible with Maxwell as long as they are built to include PTX versions of their kernels. To test that PTX JIT is working for your application, you can do the following:
5
Maxwell Compatibility Guide, Release 12.3
6
Chapter 2. Application Compatibility on Maxwell
Chapter 3. Verifying Maxwell Compatibility for Existing Applications
Chapter 4. Building Applications with Maxwell Support
When a CUDA application launches a kernel, the CUDA Runtime determines the compute capability of each GPU in the system and uses this information to automatically find the best matching cubin or PTX version of the kernel that is available. If a cubin file supporting the architecture of the target GPU is available, it is used; otherwise, the CUDA Runtime will load the PTX and JIT-compile that PTX to the GPU’s native cubin format before launching it. If neither is available, then the kernel launch will fail. The method used to build your application with either native cubin or at least PTX support for Maxwell depend on the version of the CUDA Toolkit used. The main advantages of providing native cubins are as follows:
1 Future CUDA Toolkit version might deprecate support for the Maxwell Architecture.
8
Chapter 3. Verifying Maxwell Compatibility for Existing Applications
7
3.1 Applications Using CUDA Toolkit 5.5 or Earlier . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 Applications Using CUDA Toolkit 6.0 or Later . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Maxwell Compatibility Guide for CUDA Applications The guide to building CUDA applications for GPUs based on the NVIDIA Maxwell Architecture.
Contents
1
Maxwell Compatibility Guide, Release 12.3
The first step is to check that Maxwell-compatible device code (at least PTX) is compiled in to the application. The following sections show how to accomplish this for applications built with different CUDA Toolkit versions.
3
Maxwell Compatibility Guide, Release 12.3
4
Chapter 1. About this Document
Chapter 2. Application Compatibility on Maxwell
The NVIDIA CUDA C++ compiler, nvcc, can be used to generate both architecture-specific cubin files and forward-compatible PTX versions of each kernel. Each cubin file targets a specific computecapability version and is forward-compatible only with GPU architectures of the same major version number. For example, cubin files that target compute capability 3.0 are supported on all computecapability 3.x (Kepler) devices but are not supported on compute-capability 5.x (Maxwell) devices. For this reason, to ensure forward compatibility with GPU architectures introduced after the application has been released, it is recommended that all applications include PTX versions of their kernels. Note: CUDA Runtime applications containing both cubin and PTX code for a given architecture will automatically use the cubin by default, keeping the PTX path strictly for forward-compatibility purposes. Applications that already include PTX versions of their kernels should work as-is on Maxwell-based GPUs. Applications that only support specific GPU architectures via cubin files, however, will need to be updated to provide Maxwell-compatible PTX or cubins.
4 Building Applications with Maxwell Support
9
4.1 Applications Using CUDA Toolkit 5.5 or Earlier . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2 Applications Using CUDA Toolkit 6.0 or Later . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
7
Maxwell Compatibility Guide, Release 12.3
3.2. Applications Using CUDA Toolkit 6.0 or Later
CUDA applications built using CUDA Toolkit 6.0 or Later1 are compatible with Maxwell as long as they are built to include kernels in either Maxwell-native cubin format (see Building Applications with Maxwell Support) or PTX format (see Applications Using CUDA Toolkit 5.5 or Earlier) or both.
6.3 Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
i
ii
Maxwell Compatibility Guide, Release 12.3
5 Revision History
13
6 Notices
15
6.1 Notice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2 OpenCL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16